This dataset contains forecasted weekly numbers of reported COVID-19 incident cases, incident deaths, and cumulative deaths in the United States, previously reported on COVID Data Tracker (https://covid.cdc.gov/covid-data-tracker/#datatracker-home). These forecasts were generated using mathematical models by CDC partners in the COVID-19 Forecast Hub (https://covid19forecasthub.org/doc/ensemble/). A CDC ensemble model was produced every week using the submitted models from that week at the national, and state/territory level.

This dataset is intended to mirror the observed and forecasted data, previously available for download on the CDC’s COVID Data Tracker. Mortality forecasts for both new and cumulative reported COVID-19 deaths were produced at the state and territory level and national level. Forecasts of new reported COVID-19 cases were produced at the county, state/territory, and national level. Please note that this dataset is not complete for every model, date, location or combination thereof. Specifically, county level submissions for COVID-19 incident cases were accepted, but not required, and are missing or incomplete for many models and dates. State and territory-level forecasts are more complete, but not all models submitted forecasts for all locations, dates, and targets (new reported deaths, new reported cases, and cumulative reported deaths). Forecasts for COVID-19 incident cases were discontinued in February 2022. Forecasts for COVID-19 cumulative and incident deaths were discontinued in March 2023.

This dataset includes three tables with the model-based projections and estimates as shown on CalCAT in 2025 (http://calcat.cdph.ca.gov) for California state, regions, and counties. (1) COVID-19 Nowcasts includes the R-effective estimates for COVID-19 from the different models available for the past 80 days from the archive date and the median ensemble thereof. (2) CalCAT Forecasts includes hospital census and admissions forecasts for COVID-19 and Influenza, and the corresponding ensemble metrics for a 4 week horizon from the archive date. (3) Variant Proportion Nowcasts contains the Integrated Genomic Epidemiology Dataset (IGED)-based and Terra-based estimates of COVID-19 variants circulating over the past 3 months as well as model-based predictions for the proportions of the variants of concern for dates leading up to the archive date. Prediction intervals are included when available. This dataset provides CalCAT users with programmatic access to the downloadable datasets on CalCAT. This dataset also includes a zipped file with the historical archives of the COVID-19 Nowcasts, CalCAT Forecasts and Variant Proportion Nowcasts through 2023.

Context

In partnership with the Harvard Global Health Institute, Google Cloud is releasing the COVID-19 Public Forecasts to serve as an additional resource for first responders in healthcare, the public sector, and other impacted organizations preparing for what lies ahead. These forecasts are available for free and provide a projection of COVID-19 cases, deaths, and other metrics over the next 14 days for US counties and states. For more info, see https://cloud.google.com/blog/products/ai-machine-learning/google-cloud-is-releasing-the-covid-19-public-forecasts and https://storage.googleapis.com/covid-external/COVID-19ForecastWhitePaper.pdf

Content

A projection of COVID-19 cases, deaths, and other metrics over the next 14 days for US counties and states

Acknowledgements

Released on BigQuery by Google Cloud:

https://cloud.google.com/blog/products/ai-machine-learning/google-cloud-is-releasing-the-covid-19-public-forecasts

https://pantheon.corp.google.com/marketplace/product/bigquery-public-datasets/covid19-public-forecasts

The global COVID-19 diagnosis market is projected to reach USD XXX million by 2033, with a CAGR of XX% during the forecast period 2025-2033. The market is driven by the increasing incidence of COVID-19, the rising demand for early and accurate diagnosis, and the growing adoption of molecular diagnostic tests. The market is segmented into two main types of tests: RT-PCR (Reverse Transcription Polymerase Chain Reaction) and isothermal nucleic acid amplification (INAAT). RT-PCR is the most commonly used test, as it is highly accurate and sensitive. However, it is also more expensive and time-consuming than INAAT. INAAT is a newer technology that is becoming increasingly popular, as it is faster and less expensive than RT-PCR. The market is also segmented by application, with hospitals and laboratories being the two main end-users. Hospitals are expected to account for the larger share of the market, as they are more likely to have the necessary equipment and expertise to perform COVID-19 tests. Laboratories are expected to play an increasingly important role in the market, as they are able to offer a wider range of testing services. With the global COVID-19 pandemic continuing to impact healthcare systems worldwide, the demand for accurate and reliable diagnostic tests has skyrocketed. The market for COVID-19 diagnosis has experienced significant growth, with advancements in technology and innovation driving the development of novel and efficient testing methods.

In April 2022, Tourism Economics conducted a third wave forecast study for VTC regarding the ongoing and expected impact of COVID-19 on Virginia’s travel and tourism industry. The analysis was done both at the state and regional level. Below are links to a summary report along with the complete data file containing both the statewide and regional data.

This dataset contains large files which can be used to reproduce the results in McDonald, D.J., Bien, J., Green, A., Hu, A.J., DeFries, N., Hyun, S., Oliveira, N.L., Sharpnack, J., Tang, J., Tibshirani, R., Ventura, V., Wasserman, L., and Tibshirani, R.J. “Can Auxiliary Indicators Improve COVID-19 Forecasting and Hotspot Prediction?,” Proceedings of the National Academy of Sciences, 2021. https://doi.org/10.1101/2021.06.22.21259346 Short-term forecasts of traditional streams from public health reporting (such as cases, hospitalizations, and deaths) are a key input to public health decision-making during a pandemic. Since early 2020, our research group has worked with data partners to collect, curate, and make publicly available numerous real-time COVID-19 indicators, providing multiple views of pandemic activity in the U.S. This paper studies the utility of five such indicators---derived from de-identified medical insurance claims, self-reported symptoms from online surveys, and COVID-related Google search activity---from a forecasting perspective. For each indicator, we ask whether its inclusion in an autoregressive (AR) model leads to improved predictive accuracy relative to the same model excluding it. Such an AR model, without external features, is already competitive with many top COVID-19 forecasting models in use today. Our analysis reveals that (a) inclusion of each of these five indicators improves on the overall predictive accuracy of the AR model; (b) predictive gains are in general most pronounced during times in which COVID cases are trending in "flat" or "down" directions; (c) one indicator, based on Google searches, seems to be particularly helpful during "up" trends. Complete descriptions as well as code are available from https://github.com/cmu-delphi/covidcast-pnas/ and are permanently accessible at https://doi.org/10.5281/zenodo.5639567. This material is based on work supported by gifts from Facebook, Google.org, the McCune Foundation, and Optum.

The outbreak of COVID-19, also known as novel coronavirus, has led to revised growth forecasts for global IT spending. The current forecast shows global IT industry declining by 5.1 percent in 2020 compared to the previous year. This is a further decline compared to already adjusted forecasts from April 2020. The data from the March 2020 forecast provided two possible scenarios for the impact of the coronavirus pandemic on global IT spending. In the "probable" scenario the IT spending is projected to grow by 3.7 percent compared to 2019. The"pessimistic" scenario shows a growth of 1.3 percent in 2020. The newest release now even exceeds the pessimistic scenario from that forecast. Instead of a small growth the IT market is now set to shrink in 2020.

For further information about the coronavirus (COVID-19) pandemic, please visit our dedicated Fact and Figures page.

All mfiles starting with 'Run_*' can be run on Matlab without any pre-loaded variables or data.

The three files in the root directory consist of the only input data to all program mfiles. Of which, SICM.m is the mathematical model, Input_Matched_InitialsAndParameters.mat contains initial conditions and parameter values for each day from day 50 to day 590, 30 best-fitted values for each day, ranked from the best to worst fit. The third csv file is the US cases and deaths data from CDC.

The order in which to generate the output data files is to run all Run_* mfiles in folders (1) Build_Real_Data (2) Build_Solution_Book (3) Build_Forecast_Book (4) Plot_Output

When you reach the Plot_Output folder, you have build all data files for plotting and annimation.

Run "Run_Plot_Output" will generate the first batch of output plots.

Note: It requires zero knowledge to verify the result. All one needs is the model SICM.m and the initial values and parameters provided to explore the dynamics of the model, and to analyze the fit of the model to the data, with the parameter and initial values provided from the data file Input_Matched_InitialsAndParameters.mat'. All the remaining folders contain Matlab programs to present the fit and the forecast presented in the manuscriptForecast U.S. Covid-19 Numbers by Open SIR Model with Testing'.

The dataset contains observed and 4 weeks forecast new and total weekly COVID-19 deaths at national and state level until March 9, 2023. Forecasting teams predict numbers of deaths using different types of data (e.g., COVID-19 data, demographic data, mobility data), methods, and estimates of the impacts of interventions (e.g., social distancing, use of face coverings).

Abstract INTRODUCTION: We evaluated the performance of Bayesian vector autoregressive (BVAR) and Holt’s models to forecast the weekly COVID-19 reported cases in six units of a large hospital. METHODS: Cases reported from epidemiologic weeks (EW) 12-37 were selected as the training period, and from EW 38-41 as the test period. RESULTS: The models performed well in forecasting cases within one or two weeks following the end of the time-series, but forecasts for a more distant period were inaccurate. CONCLUSIONS: Both models offered reasonable performance in very short-term forecasts for confirmed cases of COVID-19.

The dataset contains observed and 4 weeks forecast of new COVID-19 hospitalizations at national and state level. Forecasting teams predict numbers of hospitalizations using different types of data (e.g., COVID-19 data, demographic data, mobility data), methods, and estimates of the impacts of interventions (e.g., social distancing, use of face coverings).

By 2026, it is expected that sotrovimab, Glaxosmithkline and Vir's monoclonal antibody drug used for treating respiratory distress in a person infected with the COVID-19 virus, will reach a value of *** billion dollars on the worldwide pharmaceutical market. This statistic depicts the market forecast for COVID-19 antiviral and monoclonal antibody outpatient treatments worldwide from 2020 to 2026.

For more information, see the Google Cloud Blog . Developed on Google Cloud’s robust infrastructure with guidance from the Harvard Global Health Institute, the COVID-19 Public Forecasts offer a prediction of COVID-19's impact over the next 28 days. The forecasts are generated from a novel time series machine learning approach that combines AI with a robust epidemiological foundation and are trained on public data. The forecasts are maintained by Google Cloud to ensure they remain up-to-date in the changing landscape. For more detail on how the model works, see the White Paper . Forecasts are available for US state and county and Japan prefecture. US User Guide , Japan User Guide ( English and Japanese ). We encourage users who intend to make decisions in part based on these forecasts to closely review the Fairness Analysis . All bytes processed in queries against this dataset will be zeroed out making this part of the query free. Data joined with the dataset will be billed at the normal rate to prevent abuse. After September 2021, queries over these datasets will revert to the normal billing rate. This dataset is hosted in BigQuery and included in BigQuery's 1TB/mo of free tier processing. Each user receives 1TB of free BigQuery processing every month, which can be used to run queries on this public dataset. What is BigQuery?

This project consists of forecasting methods for the datasets of Covid 19 for Sachsen and Czechia, and the associated data. In a live setting it is automatically updated via a CI/CD pipeline (e.g. GitLab), and uploaded to a database that can be then accessed by a webserver backend, or a similar data consumer. This dataset has served as a basis for Where2Test website forecast dashboards for Sachsen and Czechia.

This bucket contains FAIR COVID-19 US county level forecast data

Dataset

This dataset was created by Zhuole Lu

Released under CC0: Public Domain

Contents

The Americas COVID-19 testing market was valued at around USD 5.8 Billion in 2025, backed by the continued need for diagnostic products and services amid persistent monitoring and surveillance initiatives. The global market is estimated to be over USD 10.5 Billion by 2035 at a CAGR of 6.2%.

Metric	Value
Market Size in 2025	USD 5.8 Billion
Projected Market Size in 2035	USD 10.5 Billion
CAGR (2025 to 2035)	6.2%

Country Wise Outlook

Country	CAGR (2025 to 2035)
USA	6.3%

Country	CAGR (2025 to 2035)
Canada	6.1%

Country	CAGR (2025 to 2035)
Mexico	6.2%

Country	CAGR (2025 to 2035)
Brazil	6.3%

Country	CAGR (2025 to 2035)
Argentina	6.1%

The epidemic model called SI-kJalpha is used on this website to see forecasts for new weekly and cumulative cases and deaths for all locations for which Google makes its data public. The data is noisy for some regions with decreasing cumulative values and missing values. In some cases, the forecast for a region may be lower than one of its sub-region which could be a result of less availabiltiy and more noise at the sub-region level. In that case, the sub-region data and forecast are less reliable, yet not impossible, and may point toward possible new outbreaks.

The repository contains scripts and outputs of COVID-19 forecasting developed by University of California, Santa Barbara.. The data-driven model named ACTS forecasts COVID-19 related data, e.g. confirmed cases, hospitalizations and deaths, along time. They assume that the development of the pandemic in the current region will be highly similar to another region with similar patterns a few months ago. Attention mechanism is used to compare and match such patterns and generate forecasts. Additional features such as demographic data and medical resources are leveraged to more precisely measure the similarity between regions.

Background: This study aims to estimate the total number of infected people, evaluate the effects of NPIs on the healthcare system, and predict the expected number of cases, deaths, hospitalizations due to COVID-19 in Turkey.Methods: This study was carried out according to three dimensions. In the first, the actual number of infected people was estimated. In the second, the expected total numbers of infected people, deaths, hospitalizations have been predicted in the case of no intervention. In the third, the distribution of the expected number of infected people and deaths, and ICU and non-ICU bed needs over time has been predicted via a SEIR-based simulator (TURKSAS) in four scenarios.Results: According to the number of deaths, the estimated number of infected people in Turkey on March 21 was 123,030. In the case of no intervention the expected number of infected people is 72,091,595 and deaths is 445,956, the attack rate is 88.1%, and the mortality ratio is 0.54%. The ICU bed capacity in Turkey is expected to be exceeded by 4.4-fold and non-ICU bed capacity by 3.21-fold. In the second and third scenarios compliance with NPIs makes a difference of 94,303 expected deaths. In both scenarios, the predicted peak value of occupied ICU and non-ICU beds remains below Turkey's capacity.Discussion: Predictions show that around 16 million people can be prevented from being infected and 94,000 deaths can be prevented by full compliance with the measures taken. Modeling epidemics and establishing decision support systems is an important requirement.